Deep Learning Quick Reference

Summary

In this chapter, we looked at document classification in its general form, and in the specific case of sentiment analysis. In doing so, we covered a great many NLP topics, including Bag of Word models, Vector Space models, and the relative merits of each. We also looked at using LSTMs and 1D convolutions for text analysis. We ended by training two separate document classifiers, applying everything we talked about with practical examples.

In the next chapter, we will talk about a very cool natural language model that will allow us to actually generate words, called a sequence-to-sequence model.